Situational Awareness invests $400M in Source Foundry

Leopold Aschenbrenner’s AI-focused fund raised its total backing of the chip startup to $500 million despite losses and a sharp drop in assets.

Situational Awareness, the AI-focused hedge fund founded by former OpenAI researcher Leopold Aschenbrenner, has committed another $400 million to chip startup Source Foundry. The new round brings its total investment in the company to $500 million, according to The Wall Street Journal. It is a striking move given the fund’s recent setbacks. Even after losses and a major portfolio sale, Situational Awareness is still putting serious money into AI hardware.

That matters beyond one fund or one startup. Chips sit at the center of the AI buildout, and companies developing advanced systems rely on them both for training and for deployment. A large investment in manufacturing technology is, in effect, a bet on the less visible part of the AI race.

A large wager on faster, cheaper chip production

Source Foundry was founded by Stanford researchers and is working to make chip manufacturing faster and cheaper. In a market where demand for AI computing power remains intense, that goal gives the company a clear place in the broader infrastructure stack.

The investment also shows that Situational Awareness still sees long-term value on the hardware side of AI. As IT-PUB News notes, the fund is making this bet even after public-market conditions turned against it. Rather than stepping back across the board, it is concentrating capital in a startup focused on improving how chips are made.

Aschenbrenner’s fund is under pressure but still active

The backdrop is difficult. Situational Awareness was forced to sell most of its public portfolio last month, and its assets under management reportedly fell from $20 billion to $10 billion.

The fund itself is also relatively new. Aschenbrenner launched it in 2024 without prior trading experience after working as a researcher at OpenAI. Early returns were reportedly strong, but the fund later took steep losses as AI infrastructure stocks declined.

Those losses led to a major reshaping of the portfolio. At the end of July, the fund sold most of its public holdings to Ken Griffin’s Citadel while keeping its Anthropic shares. The sale underlined how quickly the fund’s position had changed.

Even with that setback, the Source Foundry deal shows that Situational Awareness is still willing to make very large private-market investments. That could matter to investors looking for exposure to AI-related infrastructure, but it also underlines how concentrated and volatile those bets can be.

Why this investment drew attention

The size of the commitment is a big reason the deal stands out. A $400 million investment is unusually large for a startup and signals strong conviction from the backer.

It also draws attention to private funding’s role in the AI supply chain. Firms are not just chasing software models and consumer-facing products; they are also putting money into the chips and manufacturing capacity behind them.

For the broader market, the deal is another reminder that AI investment is not limited to chatbots and apps. Capital is also flowing into harder-to-see parts of the ecosystem, including chip production, where gains in speed and cost could shape how quickly AI infrastructure expands.

There is also an obvious note of risk. Situational Awareness has already taken heavy losses and seen its assets under management cut sharply, so this investment is likely to be watched as a test of whether the fund’s confidence in AI infrastructure can hold up after a rough stretch.

A fund still betting on the AI buildout

The Source Foundry investment suggests that, despite the turbulence, Aschenbrenner is still organizing the fund around the same broad thesis: AI infrastructure remains worth backing even when public-market conditions are weak.

What has changed is the pressure around that thesis. After such a sharp drop in assets, another huge private-market commitment carries more weight. For now, though, the message from Situational Awareness is clear — it is still writing very large checks on the AI buildout.